Modification of Sodium Channel Gating
by Lanthanum
Some Effects That Cannot Be Explained by Surface
Charge Theory
CLAY M. ARMSTRONG and GABRIEL COTA
ABSTRACT In clonal pituitary (GH3) cells we studied the changes in sodium
channel gating caused by substitution o f La s+ for Ca ~+ ion. Gating of sodium
channels was simplified by using intraceUular papain to remove inactivation. To
quantify La effects, we empirically fitted closing and the late phase o f opening o f
the channels with single exponentials, determined the opening (a) and closing (b)
rate, and plotted these rates as a function o f Vm (membrane voltage). The midpoint
o f the fraction o p e n - V m curve was also determined. Changing from Ca to La
shifted the curves for these three measures o f Na channel gating along the voltage
axis and changed their shape somewhat. Surface charge theory, in the form usually
presented, predicts equal shifts o f all three curves, with no change in shape. We
found, however, that the shift for each o f the measurements was different. 2 mM
La, for example, shifted opening kinetics by + 5 2 mV (i.e., 52 mV must be added to
the depolarization to make activation in 2 mM La as fast as in 2 mM Ca), the
fraction open voltage curve by + 4 2 . 5 mV, and the closing rate curve by + 28 mV.
The shift was an almost linear function o f log [La] for each o f the measures. The
main finding is that changing from 2 mM Ca to 10/zM La causes a positive shift o f
the opening rate and fraction open curves, but a negative shift o f the closing rate
curve. The opposite signs o f the two effects cannot be explained in terms o f surface
charge theory. We briefly discuss some alternatives to this theory.
INTRODUCTION
Divalent and trivalent cations have strong effects o n the gating properties o f
voltage-dependent ionic channels. These effects are usually explained in terms o f
surface charge theory, which originated f r o m a suggestion by A. F. Huxley (Frankenhaeuser and Hodgkin, 1957). A c c o r d i n g to this theory (see Hille, 1984, for a clear
Address reprint requests to Dr. Clay M. Armstrong, Department of Physiology, University of
Pennsylvania School of Medicine, Philadelphia, PA 19104-6085.
J. GEN.PHYSIOL.~)The Rockefeller UniversityPress. 0022-1295/90/12/1129/12 $2.00
Volume 96 December 1990 1129-1140
1129
Downloaded from jgp.rupress.org on July 28, 2017
From the Department of Physiology, University of Pennsylvania, Philadelphia, Pennsylvania
19104-6085; and the Department of Physiology, Biophysics, and Neurosciences, Centro de
Investigacion y de Estudios Avanzados del Instituto Politecnico Nacional, Mexico, DF 07000,
Mexico
1130
THE JOURNALOF GENERALPHYSIOLOGY.VOLUME96- 1990
METHODS
Na currents were recorded from GH3 cells (a clonal line of rat pituitary adenoma obtained
from American Type Culture Collection, Rockville, MD) using a fast whole-cell voltage clamp
with low resistance patch pipettes. The culturing and electrical recording procedures were as
described previously (Cota and Armstrong, 1989). The internal recording solution (see below),
contained in the patch pipette, was supplemented with papain (1 mg/ml) to remove fast
inactivation of Na channels (Cota and Armstrong, 1989). Current records were taken within
20 rain after the enzyme action.
The composition of the external recording solutions is given in Table I. External solutions
with La concentrations ranging from 1 to 2 mM were prepared by mixing the 0 Ca (no added
Ca) and 4 mM La solutions. The composition of the internal solution was (in mM): 100 NaF,
30 NaCl, 1 CaCl~, 10 EGTA, and 10 HEPES. EGTA and HEPES were neutralized (to pH 7.3)
with CsOH.
The high internal Na § concentration keeps inward Na current small and therefore
minimizes the voltage error associated with the access resistance. This error did not exceed 5
Downloaded from jgp.rupress.org on July 28, 2017
exposition), di- or trivalent cations bind to specific sites at the outer surface o f the
membrane, attracted, in one version o f the theory, by fixed negative charges
distributed on the m e m b r a n e surface. The adsorption o f Ca, for example, causes a
local change in the m e m b r a n e field without (directly) altering the membrane
potential measured by electrodes in the bulk solution. Since the gating apparatus o f
the channels is confined to the membrane, it cannot distinguish an electric field
change caused by altered m e m b r a n e voltage f r o m one caused by a change o f the
cation composition.
The theory is attractive, and, in many respects, successful. It is worth remembering, however, that there is no independent evidence for negative fixed charges
located appropriately near the gating apparatus o f the channel. Although a large
n u m b e r o f experiments involving many ionic species have been performed, it
remains true that such charges are inferred f r o m the results of a single type of
experiment, variation of ionic composition and observation o f the effects on gating
properties.
A prediction of the theory, in its uniform surface charge version (the version
usually considered) is that all aspects o f channel gating should be affected equally.
That is, the curves relating opening kinetics, closing kinetics, and the fraction of
open channels to Vm should all be shifted along the voltage axis by equal amounts.
Raising the divalent cation concentration in the medium, for example, should shift
all o f the curves equally to the right.
We report here on some curious effects o f La on the gating properties o f
mammalian Na channels. La is a very potent modulator o f gating activity (Takata et
al., 1966; Vogel, 1974; Hille et al., 1975), and has been called a "supercalcium"
(Takata et al., 1966). Its actions on channels are usually attributed to a surface
charge effect (Arhem, 1980; Brismar, 1980). In the course o f o u r experiments we
found a case in which the shifts o f the gating parameters are not only unequal, but
even o f opposite sign. This finding cannot be explained by surface charge theory,
and suggests that a new theory o f di- and trivalent cation action is needed. We offer
some tentative suggestions.
Modification of Sodium Channel Gating by Lanthanum
ARMSTRONG AND C O T A
1131
TABLE I
External Recording Solutions
Solution
Na*
Ca
La
Cl
HEPES*
2 mM Ca
4 mM La
0 Ca
155
151
158
2
---
-4
--
154
158
153
10
10
10
*Concentrations are in raM.
tThe pH of the solutionswas adjusted to 7.3 with NaOH, yieldingthe total Na concentrationgiven.
mV in most experiments. A disadvantage is that current is small near 0 mV, which in some
cases is in the middle of the activation range of Na channels.
For all experiments the holding potential was - 80 mV and the temperature was 15~
Even very low c o n c e n t r a t i o n s o f La slow a n d depress the o p e n i n g o f N a channels.
Fig. 1 c o m p a r e s o p e n i n g i n 2 m M Ca a n d in two c o n c e n t r a t i o n s o f La. Inactivation
o f the c h a n n e l s h a d b e e n previously r e m o v e d by i n t e r n a l papain, m a k i n g it possible
to study activation (opening) i n isolation. T h e u p p e r left p a n e l shows IN, (Na c u r r e n t )
traces r e c o r d e d at - 3 0 mV, first in 2 m M Ca a n d t h e n in 10 # M La (the La solutions
h a d n o a d d e d Ca). T h e c u r r e n t activates d u r i n g the 10-ms pulse to - 3 0 mV, a n d is
A
2 mM Ca vs. 10 ,uM La
-30
B
2 mM Ca vs. 2 mM La
-30
",~:...~
"----
..._
z.....~~ -
if,
---
~K
- ..........
C;:.~
"-'.-
/
-'9
f
,
i
f
.'
La
;f
; +30
0
./TP"
f 2":
? :
[ :
3 nA
!
.....
: #"
,aTft
i
.:.i
4 ms
~i
"i
.ILa
7
i
FIGURE 1. Opening and closing of Na channels in Ca and La: IN~ was recorded from
papain-treated GH3 cells, and is shown for steps from - 8 0 to - 3 0 and +30 mV. ENa was
- 10 mV. (A) Substituting 10/,M La for 2 mM Ca slowed opening at - 3 0 and +30 mV, and
slowed closing at - 8 0 mV. (B) I~, at - 3 0 mV was completely suppressed by 2 mM La, and
strongly slowed in its development at + 30 inV. O n stepping back to - 8 0 mV, the current tail
was smaller than expected, and fast. Experiment Se078G#2. Electrode resistance 0.6 Mr/.
Downloaded from jgp.rupress.org on July 28, 2017
RESULTS
1132
THE JOURNALOF GENERALPHYSIOLOGY.VOLUME96 91990
inward. O n stepping back to 80 mV (the holding potential) the current magnitude
j u m p s because o f the increased driving force for Na § ions, and then decays as the
channel gates close, with a time course that can be approximated by a single
exponential.
Current activation is distinctly slower in 10 #M La, and the final level o f current at
the end o f the step is smaller. There are two significant points regarding the current
tails. First, although current at pulse end is smaller in La, the initial amplitude o f the
tails in the two solutions is the same. As described below, this is because 10 ttM La
blocks the channels less than does 2 mM Ca. Second, the tail in La is a bit slower.
Thus changing f r o m 2 mM Ca to 10 #M La slows both activation and deactivation o f
the channels. This is the major finding in the paper.
At + 3 0 mV current is outward (ENa is near zero; see Methods). Activation o f the
channels is much faster than at - 3 0 mV, as expected, but is still significantly slower
in La than in Ca. Final current amplitude is slightly higher in La. O n stepping back
to - 8 0 mV, the current tail in La is bigger than in Ca, and its time course is slower.
The ratio o f the tail amplitudes is larger than expected f r o m the relative amplitudes
at pulse end, again because there is less channel blocking in 10 #M La.
2 mM La suppresses all channel opening at - 3 0 mV (Fig. 1 B). At + 30 mV the
channels activate very slowly, and the final level o f current is smaller. On stepping to
- 80 mV the initial tail current is reduced, and current decays more rapidly than in 2
mM Ca. Thus 2 mM La slows opening kinetics and speeds closing. The tail in La is
smaller than expected from the current at the end of the activating pulse because 2
mM La has a greater tendency to block Na channels than does 2 mM Ca (see below).
-
Perhaps the most familiar finding when increasing divalent concentration is a shift of
the channel activation, or fraction open, curve to the right along the voltage axis. In
terms o f surface charge, this is said to result f r o m a change of the m e m b r a n e field,
caused by binding or accumulation o f the di- or trivalent cation near the outer
m e m b r a n e surface. La exerts this effect very strongly, as shown in Fig. 2. The curves
in the figure, which we will call the ' o p e n ' or fraction open curves, are plots of the
initial amplitude of the tail current (at - 8 0 mV) as a function of Vm during the
10-ms activating pulse that preceded the tail measurement. The activating voltage is
given on the abscissa. Tail current amplitude is directly proportional to the n u m b e r
of channels that are open at the end of the activating pulse. For example, the tail
after a pulse to + 40 mV, which opens almost all of the channels, is larger than the
one after a pulse to - 4 0 mV, which opens few channels. The absolute amplitude of
the tail is affected by blocking o f the channels by Ca or La as described below.
Nonetheless, the proportionality between tail amplitude and the n u m b e r o f open
channels holds if the voltage is always returned to the same level ( - 80 mV).
In almost all cases 'open' curves saturate as voltage is increased, and in some cases
decline slightly above + 20 or 30 inV. All curves have been normalized relative to
their maximum amplitude. The tail method is the most reliable way to measure the
open fraction, but results from another method, dividing t h e / - V curve by the open
channel I-V curve (see below), gave almost identical results with regard to both the
size o f shifts and the steepness of the curves.
Downloaded from jgp.rupress.org on July 28, 2017
La Shifts the Activation-Voltage Curve to the Right
ARMSTRONGAND COTA Modification of Sodium Channel Gating by Lanthanum
1133
"Open
........... ":r,,;::=-.-"'-'4
.......i~::::;i:. . . . l'~-----"IV
, m
~..-"~
<>
[]
2mMCa
D ...."[]
.."
"41'
."
9
[]
i
="
" ,i
9 I0
[]
92 mM La
La
#M
<;.
111.
-9
,=
!
-80
"" -6~
v
.,(.-"
_~
"
:
-2"0
:
!
20
:
!
40
!
60
Vm, mV
0pen
T
..~--~.r:|
.'"
..'•
2 mM Ca []
......m-y.:$::"l' . . . . . . . . . . . .
,-l~
.."
[]
...."~ .."*
t0.8. "
'"
9 50
II La
t"
9
0.5 mM La
,tiM ~."
.T
l~ O " 4
=~ ."
="
..
-80
:
= ......
-60
..
-40
,9 ..
-
I 0"2
,
,
-20
,
20
,
,
40
,
,
60
Vm, mV
FIGURE 2. The fraction o f o p e n channels as a function o f Vm. The plots give the normalized
amplitude o f the tail current after 10-ms pulses to the voltage on the abscissa. Tail amplitude
is proportional to the n u m b e r o f o p e n channels. The curves were normalized relative to the
maximum amplitude in each case. The dotted lines are least-squares fits o f the formula given
in the text. (A) Curves were recorded in the o r d e r 2 m M Ca (D), 10 #M La, 2 mM La, and 2
mM Ca (r The fitted curve for the final 2 mM Ca determination is omitted for clarity.
Experiment Se078G#2. (B) The 2 mM Ca points are the average o f four very similar
determinations, before and after 50 #M La, and before and after 0.5 mM La. Experiment
Se068G#5 & #6.
Downloaded from jgp.rupress.org on July 28, 2017
,::"
THE JOURNALOF GENERALPHYSIOLOGY VOLUME96 91990
1134
9
Fig. 2 A shows an experiment in which recordings were made successively in 2 mM
Ca, 10 #M La, 2 mM La, and 2 mM Ca. Relative to the two curves in 2 mM Ca, the
10 #M La curve is slightly right shifted and appears to be slightly less steep. The 2
mM La curve is strongly right shifted, and is distinctly less steep than the 2 mM Ca
curves. The reversibility in this experiment, and in many others, was excellent, as can
be seen from the near identity o f the two curves in 2 mM Ca. In some cases the
kinetic effects o f La above 0.5 mM were not completely reversible.
Fig. 2 B is a plot o f experiments in two cells for which the control determinations
(before and after La exposure in the two cells) were so nearly identical that they have
been averaged and are presented as a single curve. By comparison with Fig. 2 A, it is
clear that the shift is related to the La concentration.
Empirical Quantitation of Shifts and Steepness
open = If(1 + exp
( - Z e ( V - Vl/~)/kT),
TABLE
II
Midpoints and Steepness o f Fitted Open - V,,, curoes
Solution
No. of
observations
Vi/.~ C a
Vi/2 La
Shift
raV
mV
2 mM Ca
5 #M L a
10 #M L a
50 ~M L a
0.5 mM L a
2 mM L a
4 mM La
14
1
5
1
1
1
1
- 3 4 . 0 • 3.7*
-32
- 3 5 . 4 • 3.6 '~
-33.5
-31.8
-38
-29
-33.5
- 3 3 . 0 • 4.32
-18
+1.5
+4.5
+13.5
Zc~
Zl~
mV
e
e
-1.5
+ 2 . 2 • 2.32
+15.5
+33.3
+42.5
+42.5
5.4 • 0.6
5.6
5.5 • 0.6 ~
6
5.3
5.3
4.8
4.7
4.6 • 0.4 ~
4.2
3.4
2.8
3.5
*Average a n d s t a n d a r d deviation.
where Z (the steepness factor), 1/1/2 (the voltage where the fraction open was half
maximum), and the maximum fraction open were the three parameters o f the fit.
The dotted lines are the calculated curves. In general the fits are adequate, but in
most cases the fitted curve is not sharp enough in the region where the channels are
just beginning to activate. This suggests that Z is an underestimate of the charge
movement required to open a channel.
The .results from the fittings are given in Table II. In 2 mM Ca the midpoint on
the average is - 3 4 . 3 mV and the average valence is 5.5 e (electronic charges). As
noted in the Discussion, the steepness in our papain-treated cells is much greater
than in cells with inactivation intact.
In 5/~M La there is a small left shift o f - 1.5 mV. At all higher concentrations the
shift is positive, and increases from 2 mV in 10 tzM La to 42.5 mV in 2 and 4 mM La.
At all concentrations La caused a measurable drop in the steepness o f the curves.
Downloaded from jgp.rupress.org on July 28, 2017
The steepness and midpoint of the activation curves were quantified empirically by a
least-squares fitting of each curve to the formula
ARMSTRONGAND COTA Modifi~ion of Sodium Channel Gat/ng by Lanthanum
1135
T h e a p p a r e n t valence decreased f r o m 5.4 e to 4.6 e o n changing f r o m 2 m M Ca to
10 ~tM La. T h e largest d r o p was seen with 2 m M La, f r o m 5.3 e to 2.8 e.
Quamitation of the Kinetic Effects of La
La, even at 5 #M, slows channel activation and alters closing kinetics as m e a s u r e d by
the c u r r e n t tails. A n empirical measure o f these effects was obtained by fitting a
single exponential to the activation o f the c u r r e n t at various voltages, a n d a n o t h e r
exponential to decay o f the c u r r e n t d u r i n g the tail. T h e rate constants f r o m these fits
are plotted in Fig. 3 for 2 m M Ca, 10 tiM La, a n d 2 m M La.
a
ms -1
I
-100
I
I
-50
5O
Vm, mV
FIGURE 3. Opening (a) and closing (b) rates as a function of voltage. Opening rate at each
voltage was empirically measured by fitting a single exponential, and the rate constant a is
given for each solution, b is the similarly determined rate constant for closing. The shift
described in the text and plotted in Fig. 4 is the horizontal displacement from the 2 mM Ca
curve, measured along the dashed line. Experiment Se078G#2.
T h e a curves to the right o f the origin give the rate constant o f opening, which
b e c o m e s larger as V goes in the positive direction. At any voltage, the rate is slower
in 10 ~M La than in 2 m M Ca, a n d m u c h slower in 2 m M La. T h e curves c a n n o t be
m a d e to superimpose by shifting along the voltage axis. Thus adding La does n o t
seem to have effects that can be perfectly mimicked by changing the voltage.
T h e closing rates are given by the b curves to the left o f the origin. Closing rate
increases as voltage goes negative. T h e curve in 2 mM La is located 9 5 - 3 0 mV to the
right o f the o n e in 2 m M Ca. I n 10 t~M La closing rates were slower than in 2 m M
Ca, and the La curve lies to the left o f the o n e for Ca. T h e curves in La d o n o t
a p p e a r to have the same shape as the curve in 2 mM Ca.
A l t h o u g h the curves d o n o t all have the same shape, it is nonetheless o f interest to
have a measure o f their relative positions. The shifts o f the La curves relative to the
Downloaded from jgp.rupress.org on July 28, 2017
2~4Ca
1136
T H E J O U R N A L OF GENERAL PHYSIOILX)Y 9 VOLUME 9 6 - 1 9 9 0
2 mM Ca curve were measured along the dashed line. These shifts are plotted as a
function o f log [La] in Fig. 4 and are described in the next section9 The shifts are
clearly an empirical measure, and the line along which they are measured was
selected arbitrarily. Study o f Fig. 3 shows, however, that none o f the conclusions of
the next section would have been much altered by choosing a line at a different level9
60.
52.
44
Open
Shift, mV
28
20
]Z.
4.
!
!
-4
-3
I
-2
!
l
-1
t
-4,
Log [La3+], H
-12
FIGURE 4. La-induced shifts of three measurable parameters of gating, opening rate (a),
closing rate (b), and the fraction of open channels (Open), are plotted as a function of log [La].
The baseline (0 shift) is the value of each parameter in 2 mM Ca. Kinetic shifts were measured
as illustrated in Fig. 3. The data points at 10 -s M La are the average of five determinations.
The shift is different for each of the three parameters. Experiments Se068G#5, Se068H#6,
Se078G#2, Mr179G#1, #5, #9.
Relative Shifts of Activation Parameters
A basic tenet o f the uniform surface charge theory is that the gating apparatus of a
channel cannot distinguish between addition o f a di- or trivalent cation and a suitably
chosen change o f the membrane potential. Thus the gating parameters in theory are
shifted along the voltage axis without change of shape by changing the concentration
o f the multivalent cation.
Downloaded from jgp.rupress.org on July 28, 2017
36
ARMSTRONGANDCOTA Modificationof Sodium Channel Gating by Lanthanum
1137
The three measures o f the activation process were quantified as described in the
preceding sections, and the shift o f each as a function o f log [La] is plotted in Fig. 4.
It is immediately clear that the shifts o f the three parameters are not equal. Most
affected are the activation kinetics: at 2 mM La the shift is ~52 inV. The shift of the
o p e n - V curve is 42.5 mV, and the 28-mV shift o f b, the closing rate, is substantially
smaller.
At 5 and 10 #M La, not only are the shifts different in size, but the b shift is o f the
opposite sign. The regular descent o f the curves f r o m right to left makes it clear that
this result is not an accident of measurement, but is the end point o f a steady trend.
In summary, the shifts of the gating parameters are not equal, and in some cases
not even o f the same sign.
INa, normal Ized
mV
~"
-80
2
2
~
l//
-4
10~MLa
-5
2'o
Na Channel Blocking by Ca and La
It was noted in discussing Fig. 1 that the tail amplitudes were not always proportional
to the current at the end o f the pulse. An explanation for this apparent discrepancy
has been in the literature for many years: the tendency of Ca ion (and presumably
other ions as well) to block Na channels (Woodhull, 1973; Taylor et al., 1976;
Yamamoto et al., 1984; Mozhayeva et al., 1985; Worley et al., 1986; Cukierrnan et
al., 1988; Nilius, 1988). The degree o f block is voltage dependent: block is m o r e
prominent at negative voltages, and, o f course, at high concentrations o f the
blocking agent.
The best macroscopic measure o f the blocking tendency is given by the o p e n
channel (or instantaneous) I-V curve. This curve is obtained by activating the
channels with a large depolarization, and then, when most of them are open,
Downloaded from jgp.rupress.org on July 28, 2017
Vm,
FIGURE 5. Open channel I-V
curves in Ca and La. Channels
were activated by depolarizing
to +60 mV for 0.5 ms, and Vm
was then stepped to a new
value, given on the abscissa.
The plotted current was obtained by fitting an exponential
to the current trace at the new
voltage and extrapolating back
to the beginning of the step.
The plots are normalized relative to the value of the current
at +20 mV. At negative Vm,
current in 2 mM La is smaller
than in 2 mM Ca, because La
blocks channels more efficiendy. Block is slight in 10 vM
La. Experiment Se078G#2.
1138
THE JOURNAL OF GENERAL PHYSIOLOGY- VOLUME 9 6 - 1990
DISCUSSION
Several of the actions of La that are reported here cannot be explained by surface
charge theory, particularly if the charges are assumed to be uniformly distributed.
1. The three measurable parameters o f gating, activation rate, deactivation rate, and
the fraction open-voltage relation, are not shifted equally along the Vm axis (Fig.
4), as the uniform surface charge theory predicts. This discrepancy may be
resolvable by invoking a nonuniform charge distribution.
2. At low [La] the shifts o f opening and closing kinetics are o f opposite sign. This
seems impossible to explain by any modification o f the surface charge theory.
3. La causes a marked decrease in the steepness o f the open-voltage curve, which is
not explained by surface charge theory.
4. La, like Ca, blocks Na channels at negative Vm, an action that lies outside of the
surface charge theory.
5. The shifts o f the fraction o p e n - V m curve on changing to various La concentrations are o f sizes that cannot be explained simply by screening o f surface charge.
We first calculated the surface charge density necessary to account for experimentally observed shifts o f the fraction o p e n - V m curve when [Ca] was altered
(data from two cells, with [Ca] from 0.2 to 50 mM), using the Grahame equation.
Using this charge density ( - 1.5 elementary charges/nm 2, which is slightly higher
than the largest density cited by Hille, 1984), the predicted shifts on changing
from 2 mM Ca to 10 #M and 4 mM La were - 6 mV (+2.4 mV observed) and
+ 28.5 mV (+42.5 mV observed). Even larger discrepancies were found for the
opening rate-V,, curve. These discrepancies could probably be partially resolved
by postulating that the charged sites specifically bind La in preference to Ca. Hille
Downloaded from jgp.rupress.org on July 28, 2017
changing the voltage and measuring the current at the new voltage, ideally before
any channel gates have closed. To improve time resolution, we extrapolated the
current to the beginning of the step after fitting an exponential to the tail.
Open channel I-V curves are shown in Fig. 5 for 2 mM Ca and for La at 10 #M
and 2mM. The curves were normalized relative to the current at + 20 mV in each
solution. The 2 mM Ca curve is slightly sublinear for voltage negative to about - 3 0
inV. In lower [Ca] the curve is more nearly linear (not shown). The curvature in 2
mM Ca is due to Ca's blocking action. Specifically, a Ca ion enters the channel when
the membrane field is negative, and, owing to binding or limited mobility in the
channel, interferes with conduction until it finally emerges from the channel's inner
end.
From the curves it seems that 2 mM La is more effective at blocking the channels
than is 2 mM Ca, since the current is smaller at all negative voltages. This reflects
either a superior ability o f La to enter the channel, or a lower mobility/greater
tendency to bind once it does enter, or both. For both ions at 2 mM, it seems likely
that the blocking process reaches equilibrium in a time too short for us to resolve
(see Discussion).
When [La] is 5 or 10/zM, the points lie on an almost straight line. Since there is a
pronounced tendency for La to block (as in the 2 mM La curve), this can only be the
result o f the lower rate o f entry o f La into the channels at the lower concentration.
ARMSTRONGANDCOTA Modification of Sodium Channel Gating by Lanthanum
1139
This work was supported by NIH grant NS-12547.
Original version received 25 July 1989 and accepted version received 25 June 1990.
Downloaded from jgp.rupress.org on July 28, 2017
et al. (1975) invoked specific binding. Their extensive data necessitated a model
with three types of surface charge, each with different binding affinities.
Of these points, one (2) is incompatible with surface charge theory in any form,
two (3 and 4) are phenomena that lie outside of surface charge theory, and the other
two (1 and 5) are quantitative discrepancies that may be within reach of amended
surface charge theory.
A first question is the relevance of these findings to divalent cations, e.g., Ca. Do
the La results suggest a general problem with surface charge theory? This question
is, of course, hard to answer, but the following points can be made. First, previous
authors (Vogel, 1974; Arhem, 1980; Brismar, 1980) have considered La to act by
modifying surface charge. Brismar did note, however, that La reduced the steepness
of the g-V curve, and pointed out that surface charge theory provides no explanation.
Second, and most important, most if not all of the actions of La are qualitatively
similar to actions of Ca. Thus increasing the concentration of either ion slows
opening, speeds closing, shifts the fraction open-Vm curve to the right, and causes
voltage-dependent block of the channels.
Finally, it is worth remembering that uniform surface charge theory cannot
explain some of the actions previously reported for divalent cations (see Hille, 1984).
Zn 2+, an agent that is usually listed among surface charge modifiers, affects the
opening kinetics of Na channels much more than the closing kinetics (Gilly and
Armstrong, 1982). Ca (in the absence of extracellular K) causes a substantial slowing
of opening kinetics and has almost no effect on closing kinetics (Armstrong and
Matteson, 1986). These findings are similar to the unequal shifts seen with La.
Similar discrepancies have been noted for the actions of protons on Na channel
gating (Shrager, 1974; Schauf, 1983; Campbell and Hahin, 1984).
To our minds, a theory that explained all of the effects of multivalent cations,
including blocking, steepness changes, and the oppositely directed shifts in low La,
would clearly be preferable. No theory with these capabilities exists at present, but
the following suggestions will be briefly noted. (a) During the opening and closing of
Na and K channels, there are large transmembrane movements of gating charge.
Multivalent ions may bind to gating charge (Gilly and Armstrong, 1982) rather than
to sites on, for example, membrane phospholipids. (b) Ca binds in the lumen of
closed K channels, and serves as a gating cofactor for these channels, which cannot
close stably in low Ca (Armstrong and Lopez-Barneo, 1987). (c) Our major
dissatisfaction with surface charge theory is that it does not account for the blocking
effects of di- and trivalent ions. All of the multivalent cations we have tested on Na
channels block the channels as well as shifting their gating behavior along the voltage
axis. In this regard it is interesting to note that there is at 1.east a qualitative
correlation between blocking by La and the speed of closing (Fig. 1). Similarly, there
is a qualitative correlation between blocking and the shift of the conductancevoltage curve to the right along the voltage axis. We are attempting to develop from
such observations a theory that relates the blocking and gating effects of multivalent
cations.
1140
THE JOURNAL OF GENERALPHYSIOLOGY.VOLUME 9 6 . 1990
REFERENCES
Downloaded from jgp.rupress.org on July 28, 2017
Arhem, P. 1980. Effects of rubidium, caesium, strontium, barium and lanthanum on ionic current
in myelinated nerve fibres from Xeno]ms loz~is. Acta Physiologica Scandinavia. 108:7-16.
Armstrong, C. M., and J. Lopez-Bameo. 1977. External calcium ions are required for potassium
channel gating in squid neurons. Stien~. 236:712-714.
Armstrong, C. M., and D. R. Matteson. 1986. The role of calcium ions in the closing of potassium
channels. Journal of General Physiology. 87:817-832.
Brismar, T. 1980. The effect of divalent and trivalent cations on the sodium permeability of
myelinated nerve fibres of Xenopus laevis. Acta Physiologica Scandinavia. 108:23-29.
Campbell, D. T., and R. Hahin. 1984. Altered sodium gating current kinetics in frog skeletal muscle
caused by low external pH.Journal of General Physiology. 84:771-788.
Cota, G., and C. M. Armstrong. 1989. Sodium channel gating in clonal pituitary cells. The
inactivation step is not voltage dependent. Journal of Genial Physiology. 94:213-232.
Cukierman, S., W. C. Zinkand, R.J. French, and B. K. Krueger. 1988. Effects of membrane surface
charge and calcium on the gating of rat brain sodium channels in planar bilayers. Journal of
General Physiology. 92:431-447.
Frankenhaeuser, B., and A. L. Hodgkin. 1957. The action of calcium on the electrical properties of
squid axons.Journal of Physiology. 137:218-244.
Gilly, W. F., and C. M. Armstrong. 1982. Slowing of sodium channel opening kinetics in squid axon
by extracellular zinc. Journal of General Physiology. 79:935-964.
Hille, B. 1984. The Ionic Channels of Excitable Membranes. Sinaner Associates, Sunderland, MA.
426 pp.
Hille, B., A. M. Woodhull, and B. I. Shapiro. 1975. Negative surface charge near sodium channels
of nerve: divalent ions, monovalent ions, and pH. Philosophical Transactions of the Royal Society of
London B. 270:301-318.
Mozhayeva, G. N., A. P. Naumov, and E. D. Nosyreva. 1985. Potential-dependent calcium blockade
of normal and aconitine-modified sodium channels in frog node of Ranvier. General Physiology
and Biophysics. 4:425-427.
Nilius, B. 1988. Calcium block of guinea-pig heart sodium channels with and without modifications
by piperazinylindole DPI 201-106. Journa/ of Physiology. 399:537-558.
Schauf, C. L. 1983. Evidence for negative gating charges in Myxicola axons. BiophysicalJournal
42:225-231.
Shrager, P. 1974. Ionic conductance changes in voltage clamped crayfish axons at low pH. Journal
of General Physiology. 64:666-690.
Takata, M., W. F. Pickard, J. Y. Lettvin, and J. W. Moore. 1966. Ionic conductance changes in
lobster axon membrane when lanthanum is substituted for calcium. Journal of General Physiology.
50:461-471.
Taylor, R. E., C. M. Armstrong, and F. BenzaniUa. 1976. Block of sodium channels by external
calcium ions. BiophysicalJournal. 16:27a. (Abstr.)
Vogel, W. 1974. Calcium and lanthanum effects at the nodal membrane. Pfliigers Archiv. 350:2539.
Woodhull, A. M. 1973. Ionic blockage of sodium channels in nerve. Journal of General Physiology.
61:687-708.
Worley, J. F., III, R.J. French, and B. K. Kreuger. 1986. Trimethyoxonium modification of single
batrachotoxin-activated sodium channels in planar bilaryers. Changes in unit conductance and in
block by saxitoxin and calcium. Journal of General Physiology. 87:327-349.
Yamamoto, D., J. z. Yeh, and T. Narahashi. 1984. Voltage-dependent block of normal and
tetramethrin-modified single sodium channels. BiophysicalJournal. 45:337-344.